CellCnn - Representation Learning for detection of disease-associated cell subsets
CellCnn is a convolutional neural network originally adapted to process high-dimensional single-cell measurements. It can be used to detect phenotype-associated cell subpopulations from heterogeneous single-cell resolved (e.g. mass cytometry) samples.
The CellCnn software is available on Github: https://github.com/eiriniar/CellCnn
You can access the datasets analyzed in CellCnn examples here:
These datasets have been originally published in the following studies:
- Bodenmiller, B. et al. Multiplexed mass cytometry profiling of cellular states perturbed by small-molecule regulators. Nat. Biotechnol. (2012).
- Amir, E.-A. D.et al. viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia. Nat. Biotechnol. (2013).
- Levine, J. H.et al. Data-Driven Phenotypic Dissection of AML Reveals Progenitor-like Cells that Correlate with Prognosis. Cell (2015).
- Horowitz, A. et al. Genetic and environmental determinants of human NK cell diversity revealed by mass cytometry. Sci. Transl. Med. 5 (2013).